Microsoft CEO Satya Nadella addressed internal AI adoption practices during a live recording of The New York Times Hard Fork podcast, urging employees to prioritize efficiency over token volume. Nadella acknowledged that tokenmaxxing, or maximizing AI data consumption, is widespread and addictive within the technology sector. However, he emphasized that Microsoft must transition from novelty-driven usage to purposeful application. The directive responds to a broader industry pivot as enterprise AI bills escalate. Nadella advised staff to reserve frontier AI models exclusively for complex, high-value tasks rather than routine operations. He highlighted Microsoft Copilot auto mode as an implementation of this strategy, noting that the system automatically routes requests to the most cost-effective and appropriate model. The goal is to align AI utilization with measurable business outcomes rather than engaging in unchecked computational consumption. Nadella also shared a recent example of leveraging AI for practical workflow automation, describing how he developed a code-update tool that monitors workplace discussions, generates maintenance plans, and applies patches without direct managerial oversight. This operational shift aligns with Microsoft's ongoing organizational restructuring to maintain competitive agility. In October, Nadella appointed a new CEO for Microsoft's commercial business, effectively redistributing executive responsibilities to accelerate AI integration. The move followed the November appointment of a dedicated AI advisor tasked with overhauling the company's business model for the AI era. These leadership changes reflect a broader effort to modernize the 220,000-person enterprise and compete with more nimble, AI-native organizations. During the podcast taping, Hard Fork cohost Kevin Roose presented Nadella with a vintage Microsoft Advanced AI Research T-shirt, a rare artifact from 2023 when Microsoft was preparing to establish a research lab for OpenAI staff following Sam Altman's temporary removal. The project was canceled after Altman's reinstatement, making the garment a historical footnote of early industry turbulence. Nadella accepted the item amid laughter, underscoring the evolving relationship between major technology platforms and their ecosystem partners. Microsoft's current trajectory signals a maturation phase for enterprise AI deployment. Leadership is explicitly steering the company away from unoptimized scale toward targeted, economically sustainable integration. By enforcing model-tier differentiation and streamlining executive focus, Nadella is positioning Microsoft to capture AI-driven productivity gains while containing infrastructure expenditures. The directive establishes a new operational standard for internal AI usage, emphasizing measurable output, cost discipline, and strategic alignment over raw computational volume.